| description abstract | This study examines a Venturi flume’s volumetric oxygen transfer coefficient (VFKLa) by analyzing performance metrics and graphical assessment from conducting experiments under diverse conditions. Input parameters, including discharge per unit width (q), throat width (W), throat length (F), and gauge readings (Ha,Hb), were examined. Several machine learning (ML) models, such as the artificial neural network (ANN), stacked ensemble (SE), extreme gradient boosting (XGBoost), distributed random forest (DRF), generalized linear model (GLM), and adaptive neuro-fuzzy inference system (ANFIS), were compared against empirical models like multiple variable linear regression (MVLR), multiple variable nonlinear regression (MVNLR), and the existing empirical relationship. The SE model consistently outperformed all other models in predicting both dimensional and nondimensional data sets. It achieved the highest coefficient of determination (R2) and Nash-Sutcliffe efficiency (NSE) scores: 0.9109 and 0.9095 for dimensional data sets and 0.8553 and 0.8474 for nondimensional data sets, respectively. Additionally, the SE model demonstrated the lowest root-mean-square error (RMSE) and mean absolute error (MAE) values, with 0.0125 and 0.0001 for dimensional data sets and 0.0253 and 0.0004 for nondimensional data sets. Uncertainty analysis highlighted the SE model’s robustness, showing the smallest uncertainty bands (0.0493 for dimensional, 0.0991 for nondimensional data sets) compared to the other proposed models. Sensitivity analysis identified discharge (q) and Reynolds number (R) as the most influential variables in dimensional and nondimensional data sets, respectively, confirmed by correlation studies. One-way ANOVA confirmed no significant differences between actual and ML-predicted values but highlighted substantial variability in the existing empirical relation. These findings offer insights for optimizing oxygen transfer in Venturi flumes. Venturi flumes offer versatile applications across industries, agriculture, and environmental sectors. They are primarily used to monitor flow precisely in industrial processes, water conveyance systems, and agricultural operations, which guarantees effective management and operation. In water treatment facilities, Venturi flumes play a pivotal role by aerating water streams, thereby enhancing oxygenation crucial for maintaining water quality and sustaining aquatic life. Additionally, in wastewater treatment processes, these flumes facilitate the introduction of air into wastewater streams, promoting the growth of aerobic bacteria to decompose organic contaminants. In aquaculture systems, Venturi flumes contribute to oxygenating fish tanks and ponds, fostering optimal conditions for aquatic organisms’ growth and health. Furthermore, Venturi flumes find applications in environmental monitoring for measuring water flow in natural water bodies, aiding in water resource assessment and ecosystem studies. Overall, Venturi flumes play a fundamental role in various sectors by providing accurate flow measurement and facilitating processes essential for water treatment, environmental management, and sustainable development. | |
